HIP
Increased incidence of femoral fractures in small femurs and women undergoing uncemented total hip arthroplasty – why? M. P. Bonnin, C. C. Neto, T. Aitsiselmi, C. G. Murphy, N. Bossard, S. Roche From Centre Orthopédique Santy, Lyon, France
The aim of this study was to investigate the relationship between the geometry of the proximal femur and the incidence of intra-operative fracture during uncemented total hip arthroplasty (THA). We studied the pre-operative CT scans of 100 patients undergoing THA with an uncemented femoral component. We measured the anteroposterior and mediolateral dimensions at the level of division of the femoral neck to calculate the aspect ratio of the femur. Wide variations in the shape of the femur were observed, from round, to very narrow elliptic. The femurs of women were narrower than those of men (p < 0.0001) and small femurs were also narrower than large ones. Patients with an intra-operative fracture of the calcar had smaller and narrower femurs than those without a fracture (p < 0.05) and the implanted Corail stems were smaller in those with a fracture (mean size 9 vs 12, p < 0.0001). The variability of the shape of the femoral neck at the level of division contributes to the understanding of the causation of intra-operative fractures in uncemented THA. Cite this article: Bone Joint J 2015;97-B:741–8.
M. P. Bonnin, MD, Orthopaedic Surgeon C. C. Neto, MD, Orthopaedic Surgeon T. Aitsiselmi, MD, Orthopaedic Surgeon C. G. Murphy, MD, Orthopaedic Surgeon Centre Orthopédique Santy, 24 Av Paul Santy, 69008 Lyon, France. N. Bossard, MD, PhD, Statistician S. Roche, MSc, Statistician Hospices Civils de Lyon, F-69003 Lyon, France. Correspondence should be sent to Dr M. P. Bonnin; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B6. 35022 $2.00 Bone Joint J 2015;97-B:741–8. Received 25 August 2014; Accepted after revision 21 January 2015
VOL. 97-B, No. 6, JUNE 2015
The rate of intra-operative femoral fractures1-4 during primary uncemented total hip arthroplasty (THA) ranges from 0.1%5 to 11%.6 Most fractures originate in the area of the calcar, and may extend distally.1,7-9 Although many factors contribute to the cause of these fractures, including the design of the stem and the surgical technique, they are ultimately induced by a mismatch either between the broach or the implant and the bone. This mismatch occurs predominantly at either the area of division of the femoral neck or in the metaphysis.8 The risk of intra-operative fracture with uncemented components has been reported to be higher in women10-13 and in patients with dysplastic femurs,14 but there has been no previous attempt to describe the influence of the geometry or size of the femur.6,11 Using data from the National Joint Registry for England and Wales, Jameson et al15 also reported a higher rate of revision for small sizes of stem with the Corail hip prosthesis (DePuy, Warsaw, Indiana) (size 8 to 10), at a maximum followup of 7.5 years. They observed that 10.5% of all revisions were due to fractures of the femur, and suggested that “patients requiring smaller stems may be less suitable for cementless THR.”15 We hypothesised that firstly, smaller femurs have a proportionate, or disproportionate,
geometrical difference to large femurs at the level of the division of the femoral neck and secondly, that the femurs that fracture have a different geometry to those that do not, and thirdly that femurs requiring a smaller Corail stem are more likely to fracture. To our knowledge, no morphometric analysis has been undertaken of the level that the femoral neck is divided during THA. Most geometric analyses of the proximal femur have been done in the coronal, sagittal or axial planes, which do not reflect the surgical reality of THA.6,16-18 In practice, regardless of the technique, after the femoral neck is divided, the appropriate broach and size of stem is dictated by the size and shape of the opening of the femur.16-18 This study was designed to describe the shape and size of the femur in the plane of the standard division of the femoral neck at THA and, therefore, to assess our three hypotheses.
Patients and Methods During 2012, 430 consecutive primary uncemented THAs were implanted in our centre. All patients had pre-operative CT scans of the hip as part of our pre-operative planning protocol.19 Data for all our patients were prospectively recorded in an institutional database. A subgroup of 100 patients was selected, comprising the first 50 consecutive THAs in 741
742
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Fig. 1 Coronal plane CT image of the proximal femur in the long axis of the femur and femoral neck. FO, femoral offset; NL, neck length; ML, mediolateral diameter of the femoral canal at the level of the lesser trochanter (MLLT) and 5 cm and 10 cm below the lesser trochanter (ML5 and ML10).
men and women; there were 21 right hips in men and 33 in women. Exclusion criteria included previous hip surgery, previous ipsilateral fracture of either the hip or femur, an inflammatory arthropathy or osteoarthritis (OA) secondary to Crowe III or IV dysplasia.20 The mean age was 65.4 years (24 to 85) in the men and 69.4 years (50 to 86) in the women. Pre-operatively, all patients were scanned with a 64-slice multidetector scanner (Siemens Sensation, Munich, Germany) in the supine position, with their knees fully extended and legs fixed in neutral rotation. The hip joint was scanned from the anterior inferior iliac spine to 10 cm below the lesser trochanter. The mean Dose-Length Product was 1200mGy.cm.21,22 We used image-processing software dedicated to Digital Imaging and Communications in Medicine (DICOM) images (OsiriX software Pixmeo SARL, Bernex, Switzerland) to generate threedimensional (3D) reconstructions,23 on which we digitised the deepest point of the piriform fossa and the apex of the lesser trochanter.24,25 We then created two adjusted planes. The femoral plane was modified from the coronal plane in order to include the axes of the diaphysis and the femoral neck. From this plane we measured the neck inclination angle (NI), the femoral offset (FO) and the femoral neck length (NL). We also measured the mediolateral dimensions of the endosteal canal at three levels: the apex of the lesser trochanter (MLLT) and 5 cm and 10 cm below the lesser trochanter (ML5 and ML10) (Fig. 1). We defined the canal-flare index (CFI)16,18 as the ratio between MLLT and ML10. The plane of division of the femoral neck was obtained from the axial plane, inclined by 45° around an axis orthogonal to the femoral plane and including the piriform fossa. This
Fig. 2a
Fig. 2b CT image of a) small ellipse extending laterally to the piriformis fossa and b) large ellipse extending laterally to the lateral cortex of the greater trochanter. Piriformis fossa, red arrow; D, long axis; d, short axis.
plane corresponds to the plane of the division of the femoral neck. In order to analyse the shape of the femur at the level of division, we adjusted two ellipses to match the inner cortex of the femoral neck.26,27 The ‘small ellipse’ extends laterally to the piriform fossa and the ‘large ellipse’ to the lateral cortex of the greater trochanter. From the two ellipses, we measured the long axis (D), the short axis (d) and the aspect ratio, which is defined as the ratio between them (D/d) (Fig. 2). We characterised the shape of the femoral neck at this level with the aspect ratio and calculated the surface area of both ellipses using the software described. In order to investigate the influence of the level of division of the femoral neck, we reconstructed two further planes, one 5 mm more proximal to the piriform fossa (-5 mm) and one 5 mm distal (5 mm). We measured the small and large ellipse within these planes (Fig. 3). The -5 mm plane was at the level of the tip of the greater trochanter, so we could only assess the small ellipse at this level. The 5 mm plane THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA
743
Mann–Whitney test. The relationship between the aspect ratio of the ellipses and features of the femur and age were investigated with Spearman correlation coefficients. Multiple linear regression models were used to analyse the factors influencing the aspect ratio of ellipses with four variables, including gender, NI, FO and CFI. Five methods were used to cover different assumptions and features of multiple testing, but only the Bonferroni correction31,32 was reported in the manuscript and the tables. Statistical analysis used Stata software (Stata Corp LP, College Station, Texas). All tests were two-tailed and statistical significance was considered at p < 0.05.
Fig. 3 CT image showing how measurements were made at three neck-cut levels: one at the level of the piriformis fossa (red point), one 5 mm more proximal (-5 mm plane) and another 5 mm more distal (5 mm plane).
was below the piriform fossa, thus, we could only assess the large ellipse at this point. The CT software calculated the surface area of the ellipse. In order to check the consistency in the measurement of the short and long axes we recalculated the area of the ellipse (area = D/2 × d/2 × π) and verified that it was matched with the area automated from the CT software. All operations were undertaken through a posterior approach with the patient in the lateral decubitus position. We used a compaction broaching technique28-30 and the Corail uncemented femoral component (DePuy) which ranges from size 8 to size 20. The level of division of the femoral neck and the offset were determined at pre-operative planning. The goal when broaching was to obtain primary stability, both in translation and rotation. Following the use of a calcar reamer, a collared component was used except when the high offset option was used, as this is only available as collarless. The mean size of the component was 12 (8 to 16). Post-operatively, all patients were allowed to bear weight immediately without restriction, and radiographs confirmed all of the stems were well aligned. No stems migrated during the one-year follow-up period and none have been revised to date. From our initial cohort of 430 THA, we identified all the patients with an intra-operative fracture; 12 women and one man. All underwent operation for primary OA and all fractures were Mallory Type 1 calcar fractures5 that occurred during broaching. The study had ethical approval and all patients gave informed consent. Statistical analysis. We compared the morphometric criteria of the men and women using the Student’s t-test. For comparison between the groups, we used the non-parametric VOL. 97-B, No. 6, JUNE 2015
Results The dimensions of the area at the division of the femoral neck showed marked inter-individual variation. Mean values were significantly greater in men for all measurements except NI. (Tables I and II). The aspect ratio of the surface at this level for both the short and large ellipses showed wide variation among individuals, some having a ‘round aspect’ and others a ‘narrow aspect’ (Fig. 4). The mean aspect ratio was 2.2 for the large ellipse (1.8 to 2.8) and 1.6 for the small ellipse (1.3 to 2.3). The mean aspect ratio was significantly different between men and women. ‘Round’ surfaces were more common in men, while ‘narrow’ surfaces were more common in women. In total, 12 women (24%) versus one man (2%) had a ratio superior or equal to 2.4 for the large ellipse; 9 women (18%) but no men had a ratio superior or equal to 1.9 for the small ellipse (Fig. 4). At the level of the lesser trochanter, femurs that were larger in the ML dimension had smaller aspect ratios (round aspect) while those that were smaller in this dimension had greater aspect ratios (narrow aspect). The Spearman correlation between the aspect ratio of the neck and MLLT was r = -0.27, p = 0.0074 for the small ellipse and r = -0.26, p = 0.0104 for the large ellipse. Similarly, when a small size of stem was used, the section of the neck was narrower for the short and large ellipse (r = -0.32, p = 0.0012 and r = -0.40, p < 0.0001, respectively) (Fig. 5). With multiple regression analysis, the factors associated with the aspect ratio of the neck were gender, NI and CFI. Patients with coxa vara, low CFI or cylindrical femurs had rounder necks (Table III). At the -5 mm plane, the aspect ratio was greater (more narrow necks) than at the level of the piriform fossa (1.7; standard deviation (SD) 0.2 vs 1.6; SD 0.2; p < 0.0001). At the 5 mm plane the aspect ratio was lower (more round necks) than at the level of the piriform fossa (1.9; SD 0.2 vs 2.2; SD 0.2; p < 0.0001). At these levels, women had narrower necks than men (p < 0.0001). A high correlation was observed between the aspect ratios at the level of the piriform fossa and the 5 mm level (r = 0.871, p < 0.0001) as well as the -5mm level (r = 0.883, p < 0.0001) (Fig. 6, Table IV). At the level of division of the femoral neck, patients with an intra-operative fracture of the calcar had smaller and narrower femurs compared with those without a fracture
744
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Table I. Morphometric characteristics of the femurs
Neck inclination (°) Femoral offset (mm) Neck length (mm) MLLT (mm) ML5 (mm) ML10 (mm)
Series (n = 100)
Men (n = 50)
Mean (SD) 130.1 (6.1) 44 (6) 57.8 (5.8) 28.2 (3) 13.8 (2) 11.1 (2)
Mean 129.2 46.9 60.9 29.9 14.8 12.1
Range 112 to 144 31 to 59 43 to 74 21 to 36 9 to 21 7 to 20
Women (n = 50) Range 116 to 141 38 to 58 52 to 74 21 to 36 11 to 21 8 to 20
Mean 131 41 54.6 26.4 12.8 10
Range 113 to 144 31 to 59 43 to 67 21 to 31 9 to 18 7 to 13
p-value* 0.9 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001
*Student’s t-test between females and males corrected using the Bonferroni method ML, mediolateral
Table II. Dimensional characteristics of the large and the small ellipse at the level of the piriform fossa
Large ellipse
Small ellipse
Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d) Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d)
Series (n = 100)
Men (n = 50)
Mean (SD)
Range
Mean
Range
Mean
Women (n = 50) Range
p-value*
10.8 (2) 25.2 (3) 54.2 (4) 2.2 (0.2) 7.2 (2) 23.6 (3) 38.3 (4) 1.6 (0.2)
6.8 to 15.9 18 to 38 42 to 63 1.8 to 2.8 4.4 to 10.5 16 to 31 29 to 47 1.3 to 2.3
12.4 27.4 57.3 2.1 8.3 25.9 40.7 1.6
9.7 to 15.9 22 to 33 50 to 63 1.8 to 2.5 6.2 to 10.5 21 to 31 34 to 47 1.3 to 1.8
9.2 23 51 2.2 6.1 21.3 36 1.7
6.8 to 12.3 18 to 28 42 to 58 1.8 to 2.8 4.4 to 8 16 to 26 29 to 43.1 1.4 to 2.3
< 0.0001 < 0.0001 < 0.0001 0.0106 < 0.0001 < 0.0001 < 0.0001 0.0063
*Student’s t-test between females and males corrected with the Bonferroni method
(Fig. 7). In those with a fracture, the mean aspect ratio for the large ellipse was 2.4 (SD 0.3; 2 to 2.8) whereas in those without a fracture it was 2.2 (SD 0.2; 1.8 to 2.5; p = 0.004). For the small ellipse, it was respectively 1.8 (SD 0.3; 1.4 to 2.3) and 1.6 (SD 0.15; 1.3 to 2; p < 0.017). In patients with a fracture of the calcar, the mean surface area of the ellipse was significantly smaller (p < 0.0001) and the size of stem was smaller, with a mean size of 9 (8 to 13) in comparison with 12 (9 to 16) in those without a fracture (p < 0.0001).
Discussion We found a wide variation in the shape of the bone at the level of division of the femoral neck, and that women have narrower femurs. Patients with an intra-operative fracture had a narrower femur and a smaller size of stem than those without a fracture. These results confirm our hypotheses and are in accordance with the results reported by Jameson et al.15 We suggest, then, that there are a number of reasons why hips in patients requiring small sizes during THA are more likely to fail. One reason for the increased rate of intra-operative fracture is that, regardless of the design of the stem, broaches and stems increase in size at an incremental, but not proportionate, rate. The relative increase in size is greater for smaller stems than for larger stems. The surgeon’s fear of inducing a fracture in patients with narrow necks may lead to another potential cause of failure including a lack of prosthetic integration as the stem may
be undersized, or inadequate compaction of bone when broaching. A further reason for failure is instability; potentially caused by the unrecognised distal extension of an intra-operative fracture. This study has limitations. Firstly, none of the CT-created ellipses exactly match the opening of the bone through which the stem is introduced. When we first tested our hypotheses, we attempted to perform intra-operative measurements using calipers, but abandoned this due to a lack of reliability. Analysis of CT scans using the ellipse technique has been previously validated26,27 and CT-measurement permits millimetric measurements. In this study, we referenced the plane of the division of the femoral neck to the piriform fossa, which is easy to locate on CT scans and is a landmark commonly used in THA.25,33 However, the level of insertion of the stem, and, therefore, the level of division of the femoral neck, can vary depending on the local anatomy, the surgical technique and the design of the stem. Indeed in practice, this area has no real lateral limitation. The elliptical shape of the femoral neck has been described previously by Zebaze et al,34 but they measured the external contour of the neck at its middle portion. They reported a mean aspect ratio of the neck of 1.33 (1.16 to 1.56) with a more elliptical shape at its base than at its mid zone. They found that the external necks of smaller femurs were less elliptical (more round), which differs from our observations of the internal neck. Our data also show that the THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA
Number of hips
Females Males
15 15
15 11
10 10
10
6 5
4
745
7
6
5
4
4 2
1
0
0 1.3
1.4
1.5
1.6
1.7
1.8
0
1.9
2
Long/small diameter Fig. 4a
Number of hips
15
13 10
10
8
7
9
Females Males 12
8 6
5
4 2
7
6 4
2 0
0 1.9
1.8
2
2.1
2.2
2.3
2.4
1 2.5
1
0
2.6
Long/small diameter Fig. 4b Histograms showing the distribution of the aspect ratio in males and females with a) a small ellipse and b) a large ellipse. The value under each bar is the lower boundary of the class of aspect ratio value.
Aspect-ratio of the small ellipse
2.4 2.2 2 1.8 1.6 1.4 8 to 9
10
11
12
13
14 to16
Corail prosthesis size Fig. 5 Boxplots showing the mean aspect ratio of the small ellipse for each implanted stem size.
shape of the femoral neck at the level of division is narrower when it is divided more proximally. We could argue that the variations we observed in the aspect ratio of the neck are due to fluctuations in the level of division. However, the piriform fossa was a reliable landmark on the CT scans and the variation in the aspect ratio at all three levels (piriform fossa, -5 mm and 5 mm) would oppose this argument. VOL. 97-B, No. 6, JUNE 2015
In clinical practice, the shape at the level of the division cannot be evaluated pre-operatively with standard radiographs, making it impossible to predict the individual risk of fracture. We tried to correlate the aspect ratio of the femoral neck with parameters accessible pre-operatively and found that a narrow ellipse was more likely in women and those with a small femur, a varus neck or a cylindrical
746
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Table III. Multiple regression analysis: relationship between the aspect ratio of the neck–cut plane and patient characteristics (n = 100) Small ellipse Gender Neck inclination Femoral offset Canal flare index
Regression coefficient -0.117 -0.049 -0.025 -0.073
p-value 0.003 0.009 0.24 0.064
Regression coefficient -0.118 -0.041 -0.035 -0.091
p-value 0.010 0.063 0.164. 0.050
PF
PF
35
30
5 mm proximal to PF
30 Number of patients
Number of patients
40
Large ellipse
25 20 15 10 5
5 mm proximal to PF
25 20 15 10 5 0
0 1.3
1.4 1.5
1.6
1.7
1.8
1.9
2
2.1
2.2 2.3
1.5 1.6 1.7 1.8 1.9 2
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
(Aspect-ratio) Long axis / small axis
(Aspect-ratio) Long axis / small axis Fig. 6a
Fig. 6b
Histograms comparing the distribution of the aspect ratio at different neck-cut levels showing a) a small ellipse aspect ratio at the level of the piriform fossa (PF) (black) and the -5 mm plane (shaded) and b) a large ellipse aspect ratio at the level of the PF (black) and the 5 mm plane (shaded).
Table IV. Dimensional characteristics of the small and large ellipse 5 mm distal and -5 mm proximal to the plane of the piriform fossa
Large ellipse 5 mm plane
Small ellipse -5 mm plane
2
Area (cm ) Short axis (mm) Long axis (mm) Aspect ratio (D/d) Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d)
Series (n = 100)
Men (n = 50)
Mean (SD)
Women (n = 50)
Range
Mean
Range
Mean
Range
p-value*
12.4 (2) 28.2 (3) 55 (4) 1.9 (0.2) 6.5 (1) 22 (3) 37 (3) 1.7 (0.2)
7.5 to 19.6 20 to 40 44 to 64 1.5 to 2.5 3.7 to 10 16 to 30 27 to 47 1.3 to 2.2
14.5 31 59 1.9 7.7 24.5 39.6 1.6
11 to 19.6 25 to 40 53 to 64 1.5 to 2.2 5.7 to 10 19 to 30 34 to 47 1.3 to 2
11 26 53 2 5.7 20 35 1.8
7.6 to 16 20 to 35 44 to 61 1.5 to 2.4 3.7 to 7.2 16 to 24 27 to 42 1.4 to 2.2
< 0.0001 < 0.0001 < 0.0001 0.0002 < 0.0001 < 0.0001 < 0.0001 < 0.0001
* Difference between males and females (Student’st-test)
femur. In such specific situations, in order to avoid an intraoperative fracture it would be reasonable to have a tapered cemented stem available. Other parameters such as age, NL and FO were not statistically correlated to the aspect ratio of the femoral neck. The mechanism of causing a femoral fracture is not necessarily the same with all stems or in all patients.8 Our conclusions concerning the risk factors using the Corail stem may not extend to other commercially available stems. The long-term fixation of uncemented stems requires osteointegration, which is only obtained in the presence of good
primary stability with minimal micromotion.8,35 Stability can be obtained through mediolateral cortical contact in flat stems, full peripheral cortical contact in a metaphysealfit design or through a bed of compacted cancellous bone in stems such as the Corail.8,28-30 In our series, fractures were exclusively observed in patients with a narrow femur. This supports a wedge-effect mechanism of inducing a fracture, rather than a direct impact on the medial aspect of the calcar. It may be that single wedge flat prostheses with their greater aspect ratio are more suitable for small and narrow femurs.8,15,36 Berend and Lombardi11 decreased their rate THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA M. P. Bonnin and T. Aitsiselmi have received royalties from DePuy.
35
The author or one or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article.
No fracture 30
747
Fracture
Small axis (mm)
This article was primary edited by E. Moulder and first proof edited by J. Scott.
25
References
20
15
10 25
30
35
40
45
50
Long axis (mm) Fig. 7 Scatterplot showing the relationship between the long and small diameter (x and y axes) of the small ellipse in the non-fractured and fractured femurs. The circle represents non-fractured femurs, the square, fractured femurs.
of fracture of the calcar from 4.4% in 1320 hips to 0.4% in 457 hips when they switched from metaphyseal-filling to single-wedge flat stems.14 On the other hand, stabilisation of such a flat stem relies on a press-fit between the medial and lateral cortices, which can be challenging in osteoporotic femurs or in larger round femurs. In such a situation, an intra-operative fracture might occur due to direct impact on the medial cortex. This study shows a low rate of intra-operative femoral fracture with the Corail stem and does not support a change of clinical practice in patients with smaller femurs. However, it does emphasise that the variation of the shape of the femur at the level of division of the femoral neck is more important than expected and may, in part, explain the increased risk of intra-operative fracture in uncemented THA. Patients with narrow necks have a higher risk of fracture and require careful and delicate broaching. These findings are relevant when considering the design of the femoral component for THA and suggest a role for tailoring stems to gender, as well as altering the proximal aspect ratio of smaller stems to ensure proportionality. Supplementary material A description of the methods used to address the multiple testing issues and tables showing the corrected p-values are available alongside the online version of this article at www.bjj.boneandjoint.org.uk. Author contributions: M. P. Bonnin: Performed surgeries, Main investigator, Data analysis, Writing the paper. C. C. Neto: Data collection, Data analysis. T. Aitsiselmi: Performed surgeries, Data collection. C. G. Murphy: Writing the paper. N. Bossard: Data analysis, Statistical analysis. S. Roche: Statistical analysis. VOL. 97-B, No. 6, JUNE 2015
1. Berry DJ. Epidemiology: hip and knee. Orthop Clin North Am 1999;30:183–190. 2. Cameron HU. Intraoperative hip fractures: ruining your day. J Arthroplasty 2004;19(Suppl 1):99–103. 3. Schramm M, Keck F, Hohmann D, Pitto RP. Total hip arthroplasty using an uncemented femoral component with taper design: outcome at 10-year follow-up. Arch Orthop Trauma Surg 2000;120:407–412. 4. Thillemann TM, Pedersen AB, Johnsen SP, Søballe K. Inferior outcome after intraoperative femoral fracture in total hip arthroplasty: outcome in 519 patients from the Danish Hip Arthroplasty Registry. Acta Orthop 2008;79:327–334. 5. Jewett BA, Collis DK. High complication rate with anterior total hip arthroplasties on a fracture table. Clin Orthop Relat Res 2011;469:503–507. 6. Fernandez-Fernandez R, García-Elias E, Gil-Garay E. Peroperative fractures in uncemented total hip arthroplasty: results with a single design of stem implant. Int Orthop 2008;32:307–313. 7. Capello WN, D’Antonio JA, Naughton M. Periprosthetic fractures around a cementless hydroxyapatite-coated implant: a new fracture pattern is described. Clin Orthop Relat Res 2014;472:604–610. 8. Khanuja HS, Vakil JJ, Goddard MS, Mont MA. Cementless femoral fixation in total hip arthroplasty. J Bone Joint Surg [Am] 2011;93-A:500–509. 9. Mallory TH, Kraus TJ, Vaughn BK. Intraoperative femoral fractures associated with cementless total hip arthroplasty. Orthopedics 1989;12:231–239. 10. Berend ME, Smith A, Meding JB, et al. Long-term outcome and risk factors of proximal femoral fracture in uncemented and cemented total hip arthroplasty in 2551 hips. J Arthroplasty 2006;21(Suppl 2):53–59. 11. Berend KR, Lombardi AV Jr. Intraoperative femur fracture is associated with stem and instrument design in primary total hip arthroplasty. Clin Orthop Relat Res 2010;468:2377–2381. 12. Moroni A, Faldini C, Piras F, Giannini S. Risk factors for intraoperative femoral fractures during total hip replacement. Ann Chir Gynaecol 2000;89:113–118. 13. Schwartz JT Jr, Mayer JG, Engh CA. Femoral fracture during non-cemented total hip arthroplasty. J Bone Joint Surg [Am] 1989;71-A:1135–1142. 14. Berend KR, Lombardi AV Jr, Mallory TH, et al. Cerclage wires or cables for the management of intraoperative fracture associated with a cementless, tapered femoral prosthesis: results at 2 to 16 years. J Arthroplasty 2004;19 (Suppl 2):17–21. 15. Jameson SS, Baker PN, Mason J, et al. Independent predictors of failure up to 7.5 years after 35 386 single-brand cementless total hip replacements: a retrospective cohort study using National Joint Registry data. Bone Joint J 2013;95-B:747–757. 16. Husmann O, Rubin PJ, Leyvraz PF, de Roguin B, Argenson JN. Three-dimensional morphology of the proximal femur. J Arthroplasty 1997;12:444–450. 17. Maruyama M, Feinberg JR, Capello WN, D’Antonio JA. The Frank Stinchfield Award: Morphologic features of the acetabulum and femur: anteversion angle and implant positioning. Clin Orthop Relat Res 2001;393:52–65. 18. Noble PC, Kamaric E, Sugano N, et al. Three-dimensional shape of the dysplastic femur: implications for THR. Clin Orthop Relat Res 2003;417:27–40. 19. Sariali E, Mouttet A, Pasquier G, Durante E, Catone Y. Accuracy of reconstruction of the hip using computerised three-dimensional pre-operative planning and a cementless modular neck. J Bone Joint Surg [Br] 2009;91-B:333–340. 20. Crowe JF, Mani VJ, Ranawat CS. Total hip replacement in congenital dislocation and dysplasia of the hip. J Bone Joint Surg [Am] 1979;61:15–23. 21. Huda W, Ogden KM, Khorasani MR. Converting dose-length product to effective dose at CT. Radiology 2008;248:995–1003. 22. Task Group on Control of Radiation Dose in Computed Tomography. Managing patient dose in computed tomography: a report of the International Commission on Radiological Protection. Ann ICRP 2000;30:7–45. 23. Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging 2004;17:205–216. 24. Papadakis SA, Shepherd L, Babourda EC, Papadakis S. Piriform and trochanteric fossae. A drawing mismatch or a terminology error? A review. Surg Radiol Anat 2005;27:223–226. 25. Raja S, Wordsworth D, Machin JT, Burtt S, Gray J. Piriformis fossa approach in optimising femoral neck osteotomy and component positioning in hemiarthroplasty surgery for intracapsular fractured neck of femur. Ann R Coll Surg Engl 2013;95:78– 79.
748
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
26. Cobb JP, Dixon H, Dandachli W, Iranpour F. The anatomical tibial axis: reliable rotational orientation in knee replacement. J Bone Joint Surg [Br] 2008;90-B:1032– 1038. 27. Victor J, Van Doninck D, Labey L, et al. A common reference frame for describing rotation of the distal femur: a ct-based kinematic study using cadavers. J Bone Joint Surg [Br] 2009;91-B:683–690. 28. Froimson MI, Garino J, Machenaud A, Vidalain JP. Minimum 10-year results of a tapered, titanium, hydroxyapatite-coated hip stem: an independent review. J Arthroplasty 2007;22:1–7. 29. Vidalain JP. Twenty-year results of the cementless Corail stem. Int Orthop 2011;35:189–194. 30. Wangen H, Lereim P, Holm I, Gunderson R, Reikerås O. Hip arthroplasty in patients younger than 30 years: excellent ten to 16-year follow-up results with a HAcoated stem. Int Orthop 2008;32:203–208.
31. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc, B 1995;57:289–300. 32. Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Ann Stat 2001;29:1165–1188. 33. Della Valle AG, Padgett DE, Salvati EA. Preoperative planning for primary total hip arthroplasty. J Am Acad Orthop Surg 2005;13:455–462. 34. Zebaze RM, Jones A, Welsh F, Knackstedt M, Seeman E. Femoral neck shape and the spatial distribution of its mineral mass varies with its size: clinical and biomechanical implications. Bone 2005;37:243–252. 35. Engh CA, O’Connor D, Jasty M, et al. Quantification of implant micromotion, strain shielding, and bone resorption with porous-coated anatomic medullary locking femoral prostheses. Clin Orthop Relat Res 1992;285:13–29. 36. Parvizi J, Keisu KS, Hozack WJ, Sharkey PF, Rothman RH. Primary total hip arthroplasty with an uncemented femoral component: a long-term study of the Taperloc stem. J Arthroplasty 2004;19:151–156.
THE BONE & JOINT JOURNAL
Increased incidence of femoral fractures in small femurs and women undergoing uncemented total hip arthroplasty – why? M. P. Bonnin, C. C. Neto, T. Aitsiselmi, C. G. Murphy, N. Bossard, S. Roche From Centre Orthopédique Santy, Lyon, France
The aim of this study was to investigate the relationship between the geometry of the proximal femur and the incidence of intra-operative fracture during uncemented total hip arthroplasty (THA). We studied the pre-operative CT scans of 100 patients undergoing THA with an uncemented femoral component. We measured the anteroposterior and mediolateral dimensions at the level of division of the femoral neck to calculate the aspect ratio of the femur. Wide variations in the shape of the femur were observed, from round, to very narrow elliptic. The femurs of women were narrower than those of men (p < 0.0001) and small femurs were also narrower than large ones. Patients with an intra-operative fracture of the calcar had smaller and narrower femurs than those without a fracture (p < 0.05) and the implanted Corail stems were smaller in those with a fracture (mean size 9 vs 12, p < 0.0001). The variability of the shape of the femoral neck at the level of division contributes to the understanding of the causation of intra-operative fractures in uncemented THA. Cite this article: Bone Joint J 2015;97-B:741–8.
M. P. Bonnin, MD, Orthopaedic Surgeon C. C. Neto, MD, Orthopaedic Surgeon T. Aitsiselmi, MD, Orthopaedic Surgeon C. G. Murphy, MD, Orthopaedic Surgeon Centre Orthopédique Santy, 24 Av Paul Santy, 69008 Lyon, France. N. Bossard, MD, PhD, Statistician S. Roche, MSc, Statistician Hospices Civils de Lyon, F-69003 Lyon, France. Correspondence should be sent to Dr M. P. Bonnin; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B6. 35022 $2.00 Bone Joint J 2015;97-B:741–8. Received 25 August 2014; Accepted after revision 21 January 2015
VOL. 97-B, No. 6, JUNE 2015
The rate of intra-operative femoral fractures1-4 during primary uncemented total hip arthroplasty (THA) ranges from 0.1%5 to 11%.6 Most fractures originate in the area of the calcar, and may extend distally.1,7-9 Although many factors contribute to the cause of these fractures, including the design of the stem and the surgical technique, they are ultimately induced by a mismatch either between the broach or the implant and the bone. This mismatch occurs predominantly at either the area of division of the femoral neck or in the metaphysis.8 The risk of intra-operative fracture with uncemented components has been reported to be higher in women10-13 and in patients with dysplastic femurs,14 but there has been no previous attempt to describe the influence of the geometry or size of the femur.6,11 Using data from the National Joint Registry for England and Wales, Jameson et al15 also reported a higher rate of revision for small sizes of stem with the Corail hip prosthesis (DePuy, Warsaw, Indiana) (size 8 to 10), at a maximum followup of 7.5 years. They observed that 10.5% of all revisions were due to fractures of the femur, and suggested that “patients requiring smaller stems may be less suitable for cementless THR.”15 We hypothesised that firstly, smaller femurs have a proportionate, or disproportionate,
geometrical difference to large femurs at the level of the division of the femoral neck and secondly, that the femurs that fracture have a different geometry to those that do not, and thirdly that femurs requiring a smaller Corail stem are more likely to fracture. To our knowledge, no morphometric analysis has been undertaken of the level that the femoral neck is divided during THA. Most geometric analyses of the proximal femur have been done in the coronal, sagittal or axial planes, which do not reflect the surgical reality of THA.6,16-18 In practice, regardless of the technique, after the femoral neck is divided, the appropriate broach and size of stem is dictated by the size and shape of the opening of the femur.16-18 This study was designed to describe the shape and size of the femur in the plane of the standard division of the femoral neck at THA and, therefore, to assess our three hypotheses.
Patients and Methods During 2012, 430 consecutive primary uncemented THAs were implanted in our centre. All patients had pre-operative CT scans of the hip as part of our pre-operative planning protocol.19 Data for all our patients were prospectively recorded in an institutional database. A subgroup of 100 patients was selected, comprising the first 50 consecutive THAs in 741
742
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Fig. 1 Coronal plane CT image of the proximal femur in the long axis of the femur and femoral neck. FO, femoral offset; NL, neck length; ML, mediolateral diameter of the femoral canal at the level of the lesser trochanter (MLLT) and 5 cm and 10 cm below the lesser trochanter (ML5 and ML10).
men and women; there were 21 right hips in men and 33 in women. Exclusion criteria included previous hip surgery, previous ipsilateral fracture of either the hip or femur, an inflammatory arthropathy or osteoarthritis (OA) secondary to Crowe III or IV dysplasia.20 The mean age was 65.4 years (24 to 85) in the men and 69.4 years (50 to 86) in the women. Pre-operatively, all patients were scanned with a 64-slice multidetector scanner (Siemens Sensation, Munich, Germany) in the supine position, with their knees fully extended and legs fixed in neutral rotation. The hip joint was scanned from the anterior inferior iliac spine to 10 cm below the lesser trochanter. The mean Dose-Length Product was 1200mGy.cm.21,22 We used image-processing software dedicated to Digital Imaging and Communications in Medicine (DICOM) images (OsiriX software Pixmeo SARL, Bernex, Switzerland) to generate threedimensional (3D) reconstructions,23 on which we digitised the deepest point of the piriform fossa and the apex of the lesser trochanter.24,25 We then created two adjusted planes. The femoral plane was modified from the coronal plane in order to include the axes of the diaphysis and the femoral neck. From this plane we measured the neck inclination angle (NI), the femoral offset (FO) and the femoral neck length (NL). We also measured the mediolateral dimensions of the endosteal canal at three levels: the apex of the lesser trochanter (MLLT) and 5 cm and 10 cm below the lesser trochanter (ML5 and ML10) (Fig. 1). We defined the canal-flare index (CFI)16,18 as the ratio between MLLT and ML10. The plane of division of the femoral neck was obtained from the axial plane, inclined by 45° around an axis orthogonal to the femoral plane and including the piriform fossa. This
Fig. 2a
Fig. 2b CT image of a) small ellipse extending laterally to the piriformis fossa and b) large ellipse extending laterally to the lateral cortex of the greater trochanter. Piriformis fossa, red arrow; D, long axis; d, short axis.
plane corresponds to the plane of the division of the femoral neck. In order to analyse the shape of the femur at the level of division, we adjusted two ellipses to match the inner cortex of the femoral neck.26,27 The ‘small ellipse’ extends laterally to the piriform fossa and the ‘large ellipse’ to the lateral cortex of the greater trochanter. From the two ellipses, we measured the long axis (D), the short axis (d) and the aspect ratio, which is defined as the ratio between them (D/d) (Fig. 2). We characterised the shape of the femoral neck at this level with the aspect ratio and calculated the surface area of both ellipses using the software described. In order to investigate the influence of the level of division of the femoral neck, we reconstructed two further planes, one 5 mm more proximal to the piriform fossa (-5 mm) and one 5 mm distal (5 mm). We measured the small and large ellipse within these planes (Fig. 3). The -5 mm plane was at the level of the tip of the greater trochanter, so we could only assess the small ellipse at this level. The 5 mm plane THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA
743
Mann–Whitney test. The relationship between the aspect ratio of the ellipses and features of the femur and age were investigated with Spearman correlation coefficients. Multiple linear regression models were used to analyse the factors influencing the aspect ratio of ellipses with four variables, including gender, NI, FO and CFI. Five methods were used to cover different assumptions and features of multiple testing, but only the Bonferroni correction31,32 was reported in the manuscript and the tables. Statistical analysis used Stata software (Stata Corp LP, College Station, Texas). All tests were two-tailed and statistical significance was considered at p < 0.05.
Fig. 3 CT image showing how measurements were made at three neck-cut levels: one at the level of the piriformis fossa (red point), one 5 mm more proximal (-5 mm plane) and another 5 mm more distal (5 mm plane).
was below the piriform fossa, thus, we could only assess the large ellipse at this point. The CT software calculated the surface area of the ellipse. In order to check the consistency in the measurement of the short and long axes we recalculated the area of the ellipse (area = D/2 × d/2 × π) and verified that it was matched with the area automated from the CT software. All operations were undertaken through a posterior approach with the patient in the lateral decubitus position. We used a compaction broaching technique28-30 and the Corail uncemented femoral component (DePuy) which ranges from size 8 to size 20. The level of division of the femoral neck and the offset were determined at pre-operative planning. The goal when broaching was to obtain primary stability, both in translation and rotation. Following the use of a calcar reamer, a collared component was used except when the high offset option was used, as this is only available as collarless. The mean size of the component was 12 (8 to 16). Post-operatively, all patients were allowed to bear weight immediately without restriction, and radiographs confirmed all of the stems were well aligned. No stems migrated during the one-year follow-up period and none have been revised to date. From our initial cohort of 430 THA, we identified all the patients with an intra-operative fracture; 12 women and one man. All underwent operation for primary OA and all fractures were Mallory Type 1 calcar fractures5 that occurred during broaching. The study had ethical approval and all patients gave informed consent. Statistical analysis. We compared the morphometric criteria of the men and women using the Student’s t-test. For comparison between the groups, we used the non-parametric VOL. 97-B, No. 6, JUNE 2015
Results The dimensions of the area at the division of the femoral neck showed marked inter-individual variation. Mean values were significantly greater in men for all measurements except NI. (Tables I and II). The aspect ratio of the surface at this level for both the short and large ellipses showed wide variation among individuals, some having a ‘round aspect’ and others a ‘narrow aspect’ (Fig. 4). The mean aspect ratio was 2.2 for the large ellipse (1.8 to 2.8) and 1.6 for the small ellipse (1.3 to 2.3). The mean aspect ratio was significantly different between men and women. ‘Round’ surfaces were more common in men, while ‘narrow’ surfaces were more common in women. In total, 12 women (24%) versus one man (2%) had a ratio superior or equal to 2.4 for the large ellipse; 9 women (18%) but no men had a ratio superior or equal to 1.9 for the small ellipse (Fig. 4). At the level of the lesser trochanter, femurs that were larger in the ML dimension had smaller aspect ratios (round aspect) while those that were smaller in this dimension had greater aspect ratios (narrow aspect). The Spearman correlation between the aspect ratio of the neck and MLLT was r = -0.27, p = 0.0074 for the small ellipse and r = -0.26, p = 0.0104 for the large ellipse. Similarly, when a small size of stem was used, the section of the neck was narrower for the short and large ellipse (r = -0.32, p = 0.0012 and r = -0.40, p < 0.0001, respectively) (Fig. 5). With multiple regression analysis, the factors associated with the aspect ratio of the neck were gender, NI and CFI. Patients with coxa vara, low CFI or cylindrical femurs had rounder necks (Table III). At the -5 mm plane, the aspect ratio was greater (more narrow necks) than at the level of the piriform fossa (1.7; standard deviation (SD) 0.2 vs 1.6; SD 0.2; p < 0.0001). At the 5 mm plane the aspect ratio was lower (more round necks) than at the level of the piriform fossa (1.9; SD 0.2 vs 2.2; SD 0.2; p < 0.0001). At these levels, women had narrower necks than men (p < 0.0001). A high correlation was observed between the aspect ratios at the level of the piriform fossa and the 5 mm level (r = 0.871, p < 0.0001) as well as the -5mm level (r = 0.883, p < 0.0001) (Fig. 6, Table IV). At the level of division of the femoral neck, patients with an intra-operative fracture of the calcar had smaller and narrower femurs compared with those without a fracture
744
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Table I. Morphometric characteristics of the femurs
Neck inclination (°) Femoral offset (mm) Neck length (mm) MLLT (mm) ML5 (mm) ML10 (mm)
Series (n = 100)
Men (n = 50)
Mean (SD) 130.1 (6.1) 44 (6) 57.8 (5.8) 28.2 (3) 13.8 (2) 11.1 (2)
Mean 129.2 46.9 60.9 29.9 14.8 12.1
Range 112 to 144 31 to 59 43 to 74 21 to 36 9 to 21 7 to 20
Women (n = 50) Range 116 to 141 38 to 58 52 to 74 21 to 36 11 to 21 8 to 20
Mean 131 41 54.6 26.4 12.8 10
Range 113 to 144 31 to 59 43 to 67 21 to 31 9 to 18 7 to 13
p-value* 0.9 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001
*Student’s t-test between females and males corrected using the Bonferroni method ML, mediolateral
Table II. Dimensional characteristics of the large and the small ellipse at the level of the piriform fossa
Large ellipse
Small ellipse
Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d) Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d)
Series (n = 100)
Men (n = 50)
Mean (SD)
Range
Mean
Range
Mean
Women (n = 50) Range
p-value*
10.8 (2) 25.2 (3) 54.2 (4) 2.2 (0.2) 7.2 (2) 23.6 (3) 38.3 (4) 1.6 (0.2)
6.8 to 15.9 18 to 38 42 to 63 1.8 to 2.8 4.4 to 10.5 16 to 31 29 to 47 1.3 to 2.3
12.4 27.4 57.3 2.1 8.3 25.9 40.7 1.6
9.7 to 15.9 22 to 33 50 to 63 1.8 to 2.5 6.2 to 10.5 21 to 31 34 to 47 1.3 to 1.8
9.2 23 51 2.2 6.1 21.3 36 1.7
6.8 to 12.3 18 to 28 42 to 58 1.8 to 2.8 4.4 to 8 16 to 26 29 to 43.1 1.4 to 2.3
< 0.0001 < 0.0001 < 0.0001 0.0106 < 0.0001 < 0.0001 < 0.0001 0.0063
*Student’s t-test between females and males corrected with the Bonferroni method
(Fig. 7). In those with a fracture, the mean aspect ratio for the large ellipse was 2.4 (SD 0.3; 2 to 2.8) whereas in those without a fracture it was 2.2 (SD 0.2; 1.8 to 2.5; p = 0.004). For the small ellipse, it was respectively 1.8 (SD 0.3; 1.4 to 2.3) and 1.6 (SD 0.15; 1.3 to 2; p < 0.017). In patients with a fracture of the calcar, the mean surface area of the ellipse was significantly smaller (p < 0.0001) and the size of stem was smaller, with a mean size of 9 (8 to 13) in comparison with 12 (9 to 16) in those without a fracture (p < 0.0001).
Discussion We found a wide variation in the shape of the bone at the level of division of the femoral neck, and that women have narrower femurs. Patients with an intra-operative fracture had a narrower femur and a smaller size of stem than those without a fracture. These results confirm our hypotheses and are in accordance with the results reported by Jameson et al.15 We suggest, then, that there are a number of reasons why hips in patients requiring small sizes during THA are more likely to fail. One reason for the increased rate of intra-operative fracture is that, regardless of the design of the stem, broaches and stems increase in size at an incremental, but not proportionate, rate. The relative increase in size is greater for smaller stems than for larger stems. The surgeon’s fear of inducing a fracture in patients with narrow necks may lead to another potential cause of failure including a lack of prosthetic integration as the stem may
be undersized, or inadequate compaction of bone when broaching. A further reason for failure is instability; potentially caused by the unrecognised distal extension of an intra-operative fracture. This study has limitations. Firstly, none of the CT-created ellipses exactly match the opening of the bone through which the stem is introduced. When we first tested our hypotheses, we attempted to perform intra-operative measurements using calipers, but abandoned this due to a lack of reliability. Analysis of CT scans using the ellipse technique has been previously validated26,27 and CT-measurement permits millimetric measurements. In this study, we referenced the plane of the division of the femoral neck to the piriform fossa, which is easy to locate on CT scans and is a landmark commonly used in THA.25,33 However, the level of insertion of the stem, and, therefore, the level of division of the femoral neck, can vary depending on the local anatomy, the surgical technique and the design of the stem. Indeed in practice, this area has no real lateral limitation. The elliptical shape of the femoral neck has been described previously by Zebaze et al,34 but they measured the external contour of the neck at its middle portion. They reported a mean aspect ratio of the neck of 1.33 (1.16 to 1.56) with a more elliptical shape at its base than at its mid zone. They found that the external necks of smaller femurs were less elliptical (more round), which differs from our observations of the internal neck. Our data also show that the THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA
Number of hips
Females Males
15 15
15 11
10 10
10
6 5
4
745
7
6
5
4
4 2
1
0
0 1.3
1.4
1.5
1.6
1.7
1.8
0
1.9
2
Long/small diameter Fig. 4a
Number of hips
15
13 10
10
8
7
9
Females Males 12
8 6
5
4 2
7
6 4
2 0
0 1.9
1.8
2
2.1
2.2
2.3
2.4
1 2.5
1
0
2.6
Long/small diameter Fig. 4b Histograms showing the distribution of the aspect ratio in males and females with a) a small ellipse and b) a large ellipse. The value under each bar is the lower boundary of the class of aspect ratio value.
Aspect-ratio of the small ellipse
2.4 2.2 2 1.8 1.6 1.4 8 to 9
10
11
12
13
14 to16
Corail prosthesis size Fig. 5 Boxplots showing the mean aspect ratio of the small ellipse for each implanted stem size.
shape of the femoral neck at the level of division is narrower when it is divided more proximally. We could argue that the variations we observed in the aspect ratio of the neck are due to fluctuations in the level of division. However, the piriform fossa was a reliable landmark on the CT scans and the variation in the aspect ratio at all three levels (piriform fossa, -5 mm and 5 mm) would oppose this argument. VOL. 97-B, No. 6, JUNE 2015
In clinical practice, the shape at the level of the division cannot be evaluated pre-operatively with standard radiographs, making it impossible to predict the individual risk of fracture. We tried to correlate the aspect ratio of the femoral neck with parameters accessible pre-operatively and found that a narrow ellipse was more likely in women and those with a small femur, a varus neck or a cylindrical
746
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
Table III. Multiple regression analysis: relationship between the aspect ratio of the neck–cut plane and patient characteristics (n = 100) Small ellipse Gender Neck inclination Femoral offset Canal flare index
Regression coefficient -0.117 -0.049 -0.025 -0.073
p-value 0.003 0.009 0.24 0.064
Regression coefficient -0.118 -0.041 -0.035 -0.091
p-value 0.010 0.063 0.164. 0.050
PF
PF
35
30
5 mm proximal to PF
30 Number of patients
Number of patients
40
Large ellipse
25 20 15 10 5
5 mm proximal to PF
25 20 15 10 5 0
0 1.3
1.4 1.5
1.6
1.7
1.8
1.9
2
2.1
2.2 2.3
1.5 1.6 1.7 1.8 1.9 2
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
(Aspect-ratio) Long axis / small axis
(Aspect-ratio) Long axis / small axis Fig. 6a
Fig. 6b
Histograms comparing the distribution of the aspect ratio at different neck-cut levels showing a) a small ellipse aspect ratio at the level of the piriform fossa (PF) (black) and the -5 mm plane (shaded) and b) a large ellipse aspect ratio at the level of the PF (black) and the 5 mm plane (shaded).
Table IV. Dimensional characteristics of the small and large ellipse 5 mm distal and -5 mm proximal to the plane of the piriform fossa
Large ellipse 5 mm plane
Small ellipse -5 mm plane
2
Area (cm ) Short axis (mm) Long axis (mm) Aspect ratio (D/d) Area (cm2) Short axis (mm) Long axis (mm) Aspect ratio (D/d)
Series (n = 100)
Men (n = 50)
Mean (SD)
Women (n = 50)
Range
Mean
Range
Mean
Range
p-value*
12.4 (2) 28.2 (3) 55 (4) 1.9 (0.2) 6.5 (1) 22 (3) 37 (3) 1.7 (0.2)
7.5 to 19.6 20 to 40 44 to 64 1.5 to 2.5 3.7 to 10 16 to 30 27 to 47 1.3 to 2.2
14.5 31 59 1.9 7.7 24.5 39.6 1.6
11 to 19.6 25 to 40 53 to 64 1.5 to 2.2 5.7 to 10 19 to 30 34 to 47 1.3 to 2
11 26 53 2 5.7 20 35 1.8
7.6 to 16 20 to 35 44 to 61 1.5 to 2.4 3.7 to 7.2 16 to 24 27 to 42 1.4 to 2.2
< 0.0001 < 0.0001 < 0.0001 0.0002 < 0.0001 < 0.0001 < 0.0001 < 0.0001
* Difference between males and females (Student’st-test)
femur. In such specific situations, in order to avoid an intraoperative fracture it would be reasonable to have a tapered cemented stem available. Other parameters such as age, NL and FO were not statistically correlated to the aspect ratio of the femoral neck. The mechanism of causing a femoral fracture is not necessarily the same with all stems or in all patients.8 Our conclusions concerning the risk factors using the Corail stem may not extend to other commercially available stems. The long-term fixation of uncemented stems requires osteointegration, which is only obtained in the presence of good
primary stability with minimal micromotion.8,35 Stability can be obtained through mediolateral cortical contact in flat stems, full peripheral cortical contact in a metaphysealfit design or through a bed of compacted cancellous bone in stems such as the Corail.8,28-30 In our series, fractures were exclusively observed in patients with a narrow femur. This supports a wedge-effect mechanism of inducing a fracture, rather than a direct impact on the medial aspect of the calcar. It may be that single wedge flat prostheses with their greater aspect ratio are more suitable for small and narrow femurs.8,15,36 Berend and Lombardi11 decreased their rate THE BONE & JOINT JOURNAL
INCREASED INCIDENCE OF FEMORAL FRACTURES IN SMALL FEMURS AND WOMEN UNDERGOING UNCEMENTED THA M. P. Bonnin and T. Aitsiselmi have received royalties from DePuy.
35
The author or one or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article.
No fracture 30
747
Fracture
Small axis (mm)
This article was primary edited by E. Moulder and first proof edited by J. Scott.
25
References
20
15
10 25
30
35
40
45
50
Long axis (mm) Fig. 7 Scatterplot showing the relationship between the long and small diameter (x and y axes) of the small ellipse in the non-fractured and fractured femurs. The circle represents non-fractured femurs, the square, fractured femurs.
of fracture of the calcar from 4.4% in 1320 hips to 0.4% in 457 hips when they switched from metaphyseal-filling to single-wedge flat stems.14 On the other hand, stabilisation of such a flat stem relies on a press-fit between the medial and lateral cortices, which can be challenging in osteoporotic femurs or in larger round femurs. In such a situation, an intra-operative fracture might occur due to direct impact on the medial cortex. This study shows a low rate of intra-operative femoral fracture with the Corail stem and does not support a change of clinical practice in patients with smaller femurs. However, it does emphasise that the variation of the shape of the femur at the level of division of the femoral neck is more important than expected and may, in part, explain the increased risk of intra-operative fracture in uncemented THA. Patients with narrow necks have a higher risk of fracture and require careful and delicate broaching. These findings are relevant when considering the design of the femoral component for THA and suggest a role for tailoring stems to gender, as well as altering the proximal aspect ratio of smaller stems to ensure proportionality. Supplementary material A description of the methods used to address the multiple testing issues and tables showing the corrected p-values are available alongside the online version of this article at www.bjj.boneandjoint.org.uk. Author contributions: M. P. Bonnin: Performed surgeries, Main investigator, Data analysis, Writing the paper. C. C. Neto: Data collection, Data analysis. T. Aitsiselmi: Performed surgeries, Data collection. C. G. Murphy: Writing the paper. N. Bossard: Data analysis, Statistical analysis. S. Roche: Statistical analysis. VOL. 97-B, No. 6, JUNE 2015
1. Berry DJ. Epidemiology: hip and knee. Orthop Clin North Am 1999;30:183–190. 2. Cameron HU. Intraoperative hip fractures: ruining your day. J Arthroplasty 2004;19(Suppl 1):99–103. 3. Schramm M, Keck F, Hohmann D, Pitto RP. Total hip arthroplasty using an uncemented femoral component with taper design: outcome at 10-year follow-up. Arch Orthop Trauma Surg 2000;120:407–412. 4. Thillemann TM, Pedersen AB, Johnsen SP, Søballe K. Inferior outcome after intraoperative femoral fracture in total hip arthroplasty: outcome in 519 patients from the Danish Hip Arthroplasty Registry. Acta Orthop 2008;79:327–334. 5. Jewett BA, Collis DK. High complication rate with anterior total hip arthroplasties on a fracture table. Clin Orthop Relat Res 2011;469:503–507. 6. Fernandez-Fernandez R, García-Elias E, Gil-Garay E. Peroperative fractures in uncemented total hip arthroplasty: results with a single design of stem implant. Int Orthop 2008;32:307–313. 7. Capello WN, D’Antonio JA, Naughton M. Periprosthetic fractures around a cementless hydroxyapatite-coated implant: a new fracture pattern is described. Clin Orthop Relat Res 2014;472:604–610. 8. Khanuja HS, Vakil JJ, Goddard MS, Mont MA. Cementless femoral fixation in total hip arthroplasty. J Bone Joint Surg [Am] 2011;93-A:500–509. 9. Mallory TH, Kraus TJ, Vaughn BK. Intraoperative femoral fractures associated with cementless total hip arthroplasty. Orthopedics 1989;12:231–239. 10. Berend ME, Smith A, Meding JB, et al. Long-term outcome and risk factors of proximal femoral fracture in uncemented and cemented total hip arthroplasty in 2551 hips. J Arthroplasty 2006;21(Suppl 2):53–59. 11. Berend KR, Lombardi AV Jr. Intraoperative femur fracture is associated with stem and instrument design in primary total hip arthroplasty. Clin Orthop Relat Res 2010;468:2377–2381. 12. Moroni A, Faldini C, Piras F, Giannini S. Risk factors for intraoperative femoral fractures during total hip replacement. Ann Chir Gynaecol 2000;89:113–118. 13. Schwartz JT Jr, Mayer JG, Engh CA. Femoral fracture during non-cemented total hip arthroplasty. J Bone Joint Surg [Am] 1989;71-A:1135–1142. 14. Berend KR, Lombardi AV Jr, Mallory TH, et al. Cerclage wires or cables for the management of intraoperative fracture associated with a cementless, tapered femoral prosthesis: results at 2 to 16 years. J Arthroplasty 2004;19 (Suppl 2):17–21. 15. Jameson SS, Baker PN, Mason J, et al. Independent predictors of failure up to 7.5 years after 35 386 single-brand cementless total hip replacements: a retrospective cohort study using National Joint Registry data. Bone Joint J 2013;95-B:747–757. 16. Husmann O, Rubin PJ, Leyvraz PF, de Roguin B, Argenson JN. Three-dimensional morphology of the proximal femur. J Arthroplasty 1997;12:444–450. 17. Maruyama M, Feinberg JR, Capello WN, D’Antonio JA. The Frank Stinchfield Award: Morphologic features of the acetabulum and femur: anteversion angle and implant positioning. Clin Orthop Relat Res 2001;393:52–65. 18. Noble PC, Kamaric E, Sugano N, et al. Three-dimensional shape of the dysplastic femur: implications for THR. Clin Orthop Relat Res 2003;417:27–40. 19. Sariali E, Mouttet A, Pasquier G, Durante E, Catone Y. Accuracy of reconstruction of the hip using computerised three-dimensional pre-operative planning and a cementless modular neck. J Bone Joint Surg [Br] 2009;91-B:333–340. 20. Crowe JF, Mani VJ, Ranawat CS. Total hip replacement in congenital dislocation and dysplasia of the hip. J Bone Joint Surg [Am] 1979;61:15–23. 21. Huda W, Ogden KM, Khorasani MR. Converting dose-length product to effective dose at CT. Radiology 2008;248:995–1003. 22. Task Group on Control of Radiation Dose in Computed Tomography. Managing patient dose in computed tomography: a report of the International Commission on Radiological Protection. Ann ICRP 2000;30:7–45. 23. Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging 2004;17:205–216. 24. Papadakis SA, Shepherd L, Babourda EC, Papadakis S. Piriform and trochanteric fossae. A drawing mismatch or a terminology error? A review. Surg Radiol Anat 2005;27:223–226. 25. Raja S, Wordsworth D, Machin JT, Burtt S, Gray J. Piriformis fossa approach in optimising femoral neck osteotomy and component positioning in hemiarthroplasty surgery for intracapsular fractured neck of femur. Ann R Coll Surg Engl 2013;95:78– 79.
748
M. P. BONNIN, C. C. NETO, T. AITSISELMI, C. G. MURPHY, N. BOSSARD, S. ROCHE
26. Cobb JP, Dixon H, Dandachli W, Iranpour F. The anatomical tibial axis: reliable rotational orientation in knee replacement. J Bone Joint Surg [Br] 2008;90-B:1032– 1038. 27. Victor J, Van Doninck D, Labey L, et al. A common reference frame for describing rotation of the distal femur: a ct-based kinematic study using cadavers. J Bone Joint Surg [Br] 2009;91-B:683–690. 28. Froimson MI, Garino J, Machenaud A, Vidalain JP. Minimum 10-year results of a tapered, titanium, hydroxyapatite-coated hip stem: an independent review. J Arthroplasty 2007;22:1–7. 29. Vidalain JP. Twenty-year results of the cementless Corail stem. Int Orthop 2011;35:189–194. 30. Wangen H, Lereim P, Holm I, Gunderson R, Reikerås O. Hip arthroplasty in patients younger than 30 years: excellent ten to 16-year follow-up results with a HAcoated stem. Int Orthop 2008;32:203–208.
31. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc, B 1995;57:289–300. 32. Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Ann Stat 2001;29:1165–1188. 33. Della Valle AG, Padgett DE, Salvati EA. Preoperative planning for primary total hip arthroplasty. J Am Acad Orthop Surg 2005;13:455–462. 34. Zebaze RM, Jones A, Welsh F, Knackstedt M, Seeman E. Femoral neck shape and the spatial distribution of its mineral mass varies with its size: clinical and biomechanical implications. Bone 2005;37:243–252. 35. Engh CA, O’Connor D, Jasty M, et al. Quantification of implant micromotion, strain shielding, and bone resorption with porous-coated anatomic medullary locking femoral prostheses. Clin Orthop Relat Res 1992;285:13–29. 36. Parvizi J, Keisu KS, Hozack WJ, Sharkey PF, Rothman RH. Primary total hip arthroplasty with an uncemented femoral component: a long-term study of the Taperloc stem. J Arthroplasty 2004;19:151–156.
THE BONE & JOINT JOURNAL
